Quote Originally Posted by hunsnowboard
So, as I described, this table shows a three connection chain.
Example:
.............................COLUMNS.......................
Important_0........Important_1.........Important_2.
John -----------> George -----------> Rob
John -----------> George -----------> Michael
John -----------> George -----------> Peter
John -----------> George -----------> Brad
John -----------> Mark -----------> Denis
John -----------> Mark -----------> Jones
John -----------> Smith -----------> Amber
...
...
George -----------> John-----------> LLoyd
George -----------> John -----------> Daniel
Smith -----------> Amber -----------> Clark

In the real file there are more than 140.000 rows which means that there are many many connections, with names present in all three columns (column A, B C). And this is my first problem, how do I normalize this table? Could you give me an advice? How would you do it?
Your example shows only unique combinations. Is that a requirement of your data model? In other words, could you have another row which includes "John -----------> George -----------> Rob"?

If repeating combinations are expected, you might consider a table design like this:

Combination_ID...Person.......Importance
1 ---------------> John ----> 0
1 ---------------> George -> 1
1 ---------------> Rob -----> 2
2 ---------------> John ----> 0
2 ---------------> George -> 1
2 ---------------> Michael -> 2

That design would support the query you wanted for your form:
SELECT DISTINCT Importance FROM tblConnections
WHERE Person = [Enter name of Person];

Hope that helps. But I don't really understand what you're trying to accomplish.

Hans